The Australian Energy Regulator Expenditure Forecast Assessment Guidelines General discussion of draft guidelines 2 September 2013

Consultation up to the release of the final guideline 2 DateSubject 2 SeptemberGeneral discussion of the draft guidelines Mid September Release of draft economic benchmarking RIN 20 September Submissions on explanatory statement and guideline due Late September Category analysis state by state workshops Bilateral meetings / workshop on economic benchmarking data requirements Mid OctoberSubmissions on draft economic benchmarking RIN due NovemberEconomic benchmarking RIN issued Final guideline and explanatory statement released Draft category analysis RIN released

Context for today Today we are looking to get your initial reactions on: ◦the explanatory statement and guideline ◦category analysis ◦economic benchmarking We will be gathering feedback in order to determine which areas to focus on in further consultation Provide clarifications to inform stakeholder submissions 3

Agenda 9:30General guideline design, assessment approach and implementation 10:15 Category analysis 11:00Break 11:15 Economic benchmarking 12:00Summary and next steps 4

Guideline design, assessment approach, and implementation 5

Purpose of the guideline: recap Guidelines: specify the approach that the AER proposes to use to assess opex and capex set out data requirements Process: 1. At the F&A the AER decides on the application of the guideline to the NSP (data requirements) 2. NSP provides data with proposal 3. The AER can depart from the F&A, but the AER must provide reasons 6

Guideline design (1) 7 Guideline sets out the AER’s assessment approach for opex and capex ◦Plain language, rather than provisions Assessment approach is very similar to current approach Flexibility provided for AER to select best techniques or multiple techniques

Principles are not in the guideline, but are set out in explanatory statement The role of any particular technique has not been constrained Data requirements set out in appendix to guideline, but ultimately will be collected in RINs 8 Guideline design (2)

Benchmarking (economic, category) Methodology review Governance and policy review Trend analysis Cost-benefit analysis Detailed project review 9 Assessment approach: Techniques

Accuracy and reliability Robustness Transparency Parsimony Fitness for purpose 10 Assessment approach: Principles

Base step trend model Revealed costs for the base, unless we identify material inefficiencies All techniques used to assess base 11 Assessment approach: Opex (1)

Rate of change (output growth, real price growth, productivity growth, step changes) Ideally, productivity growth will be estimated using economic benchmarking 12 Assessment approach: Opex (2)

13 Assessment approach: Capex (1) All techniques applied, but will primarily review category by driver (unit cost and volume) Will be applying augex and repex models, but may need to supplement with other techniques

14 Implementation(1) Guideline and explanatory statement in place by 29 November Final RINs Economic benchmarking — Oct / Nov 2013 Category — Feb 2014 Reset — Feb annual benchmarking report Economic benchmarking Category analysis

15 Implementation: RINs(2) 2014: Economic benchmarking RIN Category analysis RIN Reset RIN (if relevant) Annual reporting RIN In 2015, we will transition to: Annual reporting RIO Reset RIN

Category analysis 16

Key elements of approach Driver/ activity based, consistent with previous reviews Much more data than requested before ◦Visibility of historic and forecast work rates/ volumes for each activity ◦Comparisons of cost per “unit” – trend and benchmarking comparisons ◦Capturing of material cost drivers and “uncontrollable” environmental factors 17

Points for discussion Level of detail ◦Labour, materials, contract, other ◦Balance between accurate, like for like comparisons vs high level/ reasonableness checks Capture of NSP idiosyncrasies Impact of service classifications ◦Inclusion of metering, public lighting, ACS Ability of NSPs to provide reliable, useful backcast data 18

Points for discussion (cont) Separation of direct/ indirect costs Addressing differences in allocations and capitalisation ◦Scope for data requirements outside of guidelines eg reconciliation to statutory accounts Issues for targeted consultation ◦Developing robust definitions - worked examples may be useful ◦Feasibility/ timing issues for NSPs ◦Consistency with economic benchmarking data 19

Economic benchmarking 20

Model specification: DNSPs (1) 21 Output specifications 1 and 2: ◦Customers ◦Capacity (kVA * Kms) [or smoothed maximum demand] ◦Interruptions Output specification 2 ◦Residential customers ◦Commercial customers ◦Small industrial ◦Large industrial ◦Interruptions

Model specification: DNSPs (2) 22 Environmental variables ◦Density — Customer density, Energy density, Demand density ◦Weather factors — Extreme heat days, Extreme cold days, Extreme wind days ◦Terrain factors — Bushfire risk, Rural proportion, Vegetation growth and encroachment, Standard vehicle access ◦Service area factor — Line length

Model specification: TNSPs (1) 23 Output specification 1: ◦System capacity (kVA*kms) ◦Entry and exit points ◦Loss of supply events ◦Aggregate unplanned outage duration Output specification 2: ◦Smoothed non-coincident maximum demand (as a capacity measure) ◦Energy delivered could be incorporated into either specification

Model specification: TNSPs (2) 24 Environmental variables: ◦Weather factors — Extreme heat days, Extreme cold days, Extreme wind days, Average wind speed ◦Terrain factors — Bushfire risk, Rural proportion, Vegetation growth and encroachment, Standard vehicle access, Altitude ◦Network characteristics — Line length, Variability of dispatch, Concentrated load distance

Model specification: Inputs 25 Input specifications ◦Nominal opex (deflated by average price index) ◦Overhead lines (MVA-kms) ◦Underground lines (MVA-kms) ◦Transformers and other (MVA) Alternative capital inputs based on nominal RAB deflated by a price index will be looked at as well

Applications 26 Models ◦MTFP, DEA (cross sectional, time series) ◦Econometric opex model Applications as set out in models (6 June workshop) ◦Overall NSP efficiency, frontier ◦Diagnosing sources of inefficiency ◦Top down opex forecast, rate of change

Next steps (1) 27 Draft RIN Consultation on draft RIN ◦Would another workshop on data be helpful?

Next steps (2) 28 Economic Insights ◦Engaged to conduct testing and validation exercise for benchmarking models in 2014 ◦To commence in February ◦Will initially review robustness of data, liaise with NSPs ◦Model and report to be provided to stakeholders

Summary and next steps 29